This proposal is designed to capitalize on youth Internet use to promote condom use, incorporating (1) a one-year formative phase to assess youth (i.e., between 18 and 24 years old) preferences for web design and to develop state-of-the- art, web-based tailored messages promoting condom use based on Social Cognitive Theory (SCT) and the Theory of Planned Behavior (TPB); and (2) a randomized controlled trial (RCT) to assess the efficacy of web-based tailored messages on condom use among youth. During the first 6 months, investigators will conduct iterative focus groups with 18-to 24-year-olds to generate, test, and modify ideas for tailored condom use messages. Simultaneous to the focus groups and for 6 months thereafter, investigators will work with expert web designers, to translate design preferences to multiple Internet web pages. In months 13 through 40 of the project, we will implement a randomized controlled trial that will take place on the Internet. Youth between the ages of 18 and 24 who attend Denver Health or Planned Parenthood clinic sites in the Denver metropolitan area will be recruited to participate in the study. They first will complete (in the clinic setting) an online, self-administered baseline assessment of condom-use outcome expectations, norms, self-efficacy and practices. Participants will then be randomized (by computer) to one of two study arms. In the first arm (control), participants will be directed to a web page with generic didactic messages about condom use. The second arm (tailored intervention) will direct participants to web pages where they will receive tailored feedback to increase positive outcome expectancies towards condom use, condom use norms, self-efficacy for condom negotiation, and self-efficacy for condom use in the form of tailored messages delivered by role models. All participants will be asked to return to the Internet site two times--at one month for a booster exposure to messages, and three months for a follow-up risk assessment. These follow-up visits can occur at any location where the participant can access the Internet. Through this investigation we will discern whether condom outcome expectancies, norms and self-efficacy are affected by tailored interactive messages and whether condom use with non-main partners is affected by tailored interactive messages.